Audyssey Manual Calibration “OCA’s REW + Audyssey Awesomeness”

[Times]

Try ticking "Play dither before sweep/silence" options to invoke the DACs before the sweep.

Just cecked that and I already had checked that option...
But I helped myself in another way:

I calculated the differences (in ms) between what the distances are in the MultEQ app measurements and the ones from REW/UMIK, then used the aligment tool, manually entered the delay in ms, made an aligned copy and used these ones going forward.

 

[OCA]

Auto-eq filters will be identical regardless of the time alignment of measurements.

 

[Times]

Thanks for confirmation. I suspected that, but wanted to be sure, just in case - and am happy that I figured something out by myself.
Let's see for how long that works...

 

[OCA]

Does the XL file you are using contain a cell for speed of sound setting? I ask because I updated that file a couple of times after the launch of the video and some older versions had minor bugs.

 

[Times]

Yes, A22 is green and says "9.4.6", D22 is for setting the speed of sound.

 

[Times]

Thanks everyone, I got it up and running with the REW / UMIK measurements and the new MultEQ App measurement as basis. I also discovered at least two errors I made the first time, while making the filters: Used "A+B" instead of "A*B" and the filter txts contained headers/remarks although the box was checked. ‍

The only thing that didn't work as intended (?) was getting the custom volume levels and distances in via CSV and browser script. I had to do that manually in the MultEQ app / AVR menu afterwards.

Sound stage and depth are amazing now, but I could use a little bit more bass and a tad less treble, so I am going to try again with another target curve... sometime.

@OCA if it's not too much of hassle, it would be great if your offer still stands regardless and you could take a brief look at my files to make sure everything is in order?


have a nice weekend,
Timo

 

[OCA]

you could take a brief look at my files

Sure thing.

 

[OCA]

Done! Just upload the attached ady file to your app and transfer as is to your PC. Audyssey should be set to Reference and everything else OFF!

 

[LeoB]

Thanks everyone, I got it up and running with the REW / UMIK measurements and the new MultEQ App measurement as basis. I also discovered at least two errors I made the first time, while making the filters: Used "A+B" instead of "A*B" and the filter txts contained headers/remarks although the box was checked. ‍

The only thing that didn't work as intended (?) was getting the custom volume levels and distances in via CSV and browser script. I had to do that manually in the MultEQ app / AVR menu afterwards.

Sound stage and depth are amazing now, but I could use a little bit more bass and a tad less treble, so I am going to try again with another target curve... sometime.

@OCA if it's not too much of hassle, it would be great if your offer still stands regardless and you could take a brief look at my files to make sure everything is in order?


have a nice weekend,
Timo

Before trying a new target curve, try average Aud multi-position responses in Rew using RMS Average function instead of Vector Average. In some cases it produces very different results.

 

[Times]

Thanks for the hint. As soon as the new AVR arrives, I am going to do the whole process from scratch again, this time with setting save points at certain steps (before averaging the measurements and before applying a house curve), to be able to try out different options without starting at 0 every time.

I've read a couple of statements regarding RMS vs. Vector average. Consensus seems to be that there's neither right nor wrong, just a slight difference between both, especially if the room reverbarates a lot. I don't think thats the case for me, though:

But since the proof of the pudding is in the eating, I'll try, measure and listen.

thanks to all,
Timo

 

[LeoB]

In my case the difference was not slight.

 

[Times]

Oh... indeed it was not.
Thanks for sharing.

---

EDIT: Just tried it with the 5-Point REW measurement of my front left speaker, both with var smoothing:

Question is which average is going to be the better basis for creating the filers and thus resulting in a "better" overall sound for my ears... I'll have to try, listen and measure.

 

[Ruspamen]

So is it better to use vector average or rms average before applying filters for eq?

 

[LeoB]

So is it better to use vector average or rms average before applying filters for eq?

From my experience - RMS average.

 

[Hayabusa]

Why would anyone possibly want do to that, though?

Just because a feature is there, does not mean it's worth a damn. Here it seems like this "directional bass" business will only cause issues.

IMO start with first priniciples - is x useful?

If you want to have stereo bass going to your left and right sub for instance? very useful.

 

[OCA]

So is it better to use vector average or rms average before applying filters for eq?

Try using the new "dB + phase average" and "calculate target curve from response" to determine the SPL level of the target to EQ. How you smooth the speaker response before auto-EQ'ing to that target curve will depend on what frequency range you want to apply correction to. For EQ below room transient frequency (200-225Hz) or subs, no smoothing or 1/48 will work nicely. For higher frequency EQ, a frequency dependent windowing (or gated windowing) along with a VAR or psychoacoustic smoothing will need to be applied (which will usually pull the HF response below target anyway deeming EQ unnecessary). I will suggest to listen and compare volume matched - before/after high frequency EQ'ed versions of your calibration. It's likely that sound will be degraded/throttled overall if you use filters at very high frequencies and ruin the additional clarity in mid/high tones your low frequency EQ has achieved.

 

[Mattknz]

I'm hoping to try this sometime in the next few days.
I'm using a Denon avr with 2 independent sub outs and have a sub plugged into each. How do I do the second sub? In my very limited quick play with REW it seems I can only send a signal via Sub out 1 .

 

[OCA]

I'm hoping to try this sometime in the next few days.
I'm using a Denon avr with 2 independent sub outs and have a sub plugged into each. How do I do the second sub? In my very limited quick play with REW it seems I can only send a signal via Sub out 1 .

That's true, computers can only send one LFE signal at a time. You will need to manually turn on/off each sub respectively and measure twice.

 

[Mattknz]

Once I have measured and created curves for both subs how do I get that info to audyssey as you can only input eq points for one sub?

 

[OCA]

Once I have measured and created curves for both subs how do I get that info to audyssey as you can only input eq points for one sub?

The scripts were updated in December slightly to handle two separate subs. Make sure you have the newer ones. Still, depending on the AVR model, you might need to create a dummy SW2.txt to upload and then delete it in the ady. If you get stuck, share your mdat and ady and I can fix it for you.